Introduction
How the human brain functions under conditions of acute hypoglycemia remains a complex question, by virtue of the potential simultaneous shifts in processes of perfusion, metabolism and changing demand. We examined this issue by measuring cerebral perfusion (CBF) and oxidative metabolism (CMRO2) in insulin induced hypoglycemia (HG) and euglycemia (EG) at rest and during sensorimotor activation in normal human subjects using magnetic resonance (MR).
Methods
Experiments were performed with 12 subjects (9 M, 3F; age 29.6±8.5). The protocol consisted of either insulin induced hypoglycemia (targeting a HG of 60 mg/dl) followed by euglycemia (n=7), or in reverse order (n=7; two volunteers studied twice), each phase approximately 1.5hrs. EG was performed with the same insulin infusion rate to match the HG phase. All MR data were acquired at least 30 min after stabilization of plasma glucose at either HG or EG levels. MR: All studies were performed with a Varian Inova whole body 4 T human MR system. IR GE images were obtained to identify the plane of interest (motor cortex). Studies acquired were perfusion (FAIR-single shot EPI, activated and non-activated), gradient echo BOLD EPI and quantitative R2' (non-activated). All physiological studies were performed with FOV 192×192, 3 mm isotropic resolution. Task activation was provoked with blocks of bilateral finger tapping. Quantitative R2' data were calculated from 16 field echoes prepared by either a 50 or 80 msec spin echo. Analysis of the BOLD data was done to identify regions of interest (105±15 pixels) in the motor and supplementary cortices. The flow and transverse relaxation data were incorporated into a model of deoxyhemoglobin concentration, oxygen extraction and consumption as described by numerous workers, R2' is the blood susceptibility induced transverse relaxation rate, R2', fM the fractional change in oxygen consumption, fP fractional change in perfusion.
Results
Stable plasma glucose levels were achieved, at 5. 78±0.03 mM, EG and 3.56±0.02 mM, HG. Basal perfusion increased significantly from 56.4±13.6 ml/100 g/min (EG) to 64.3±7.6 ml/100 g/min (HG). Perfusion further increased significantly with activation in both EG and HG and were significantly different from each other, while the delta activated flows were not different. R2' remained the same between EG and HG. The finger tapping BOLD signal decreased from EG to HG (1.79±0.60% to 1.55±0.60% p<0.02). Modeling the R2' and perfusion data together demonstrated that in hypoglycemia, flow and metabolism remained well coupled. Similarly, evaluating the BOLD data with perfusion and R2' showed that fM/fP was remained coupled with a slope of 0.44±0.16 (R=+0.87) in EG, and 0.57±0.20 (R=+0.88) in HG.
Conclusions
Although the depth of hypoglycemia achieved in the present study was mild, we were able to detect a consistent increase in perfusion with hypoglycemia. The finger tapping task activation was not associated with any larger (delta) flow values. Metabolism appears to remain coupled with flow in mild HG. Thus although mild HG induces increases in basal perfusion and metabolism, a similar increase was not seen with the elementary activation. These data demonstrate that performance declines in mild to moderate hypoglycemia occurs with flow and metabolism staying coupled and both decreasing.